Globular clusters display an anticorrelation between the fraction of the first generation of stars (N(G1)/N(tot)) and the slope of the present-day mass function of the clusters (alpha(pd)), which is particularly significant for massive clusters. In the framework of the binary-mediated collision scenario for the formation of the second-generation stars in globular clusters, we test the effect of a varying stellar initial mass function (IMF) of the G1 stars on the (N(G1)/N(tot)) - alpha(pd) anticorrelation. We use a simple collision model that has only two input parameters, the shape of the IMF of G1 stars and the fraction of G1 stars that coalesce to form second-generation stars. We show that a variable efficiency of the collision process is necessary in order to explain the (N(G1)/N(tot)) - alpha(pd) anticorrelation; however, the scatter in the anticorrelation can only be explained by variations in the IMF, and in particular by variations in the slope in the mass interval approximate to(0.1-0.5) M-circle dot. Our results indicate that in order to explain the scatter in the (N(G1)/N(tot)) - alpha(pd) relation, it is necessary to invoke variations in the slope in this mass range between approximate to -0.9 and approximate to -1.9. Interpreted in terms of a Kroupa-like broken power law, this translates into variations in the mean mass of between approximate to 0.2 and 0.55 M-circle dot. This level of variation is consistent with what is observed for young stellar clusters in the Milky Way and may reflect variations in the physical conditions of the globular cluster progenitor clouds at the time the G1 population formed or may indicate the occurrence of collisions between protostellar embryos before stars settle on the main sequence.